PSI - Issue 77

Roman Hofmann et al. / Procedia Structural Integrity 77 (2026) 237–247

241

Roman Hofmann et al. / Structural Integrity Procedia 00 (2026) 000–000

5

The scanning strategy is illustrated schematically in Figure 4a which presents a depiction of the potential arrangement of seeds in the layer. This seed distribution is then used to generate the Voronoi pattern. In accordance with the previously outlined exposure sequence optimization, the Voronoi fields are exposed zone by zone, as illustrated schematically in Figure 4b. The Voronoi approach thereby combines multiple objectives: reducing residual stresses by distributing heat input more uniformly, adapting hatch orientation and parameters to local geometry, and enabling the targeted tailoring of material properties. In principle, each Voronoi field could also incorporate additional reordering schemes, such as the index- or time-based strategies described above, further increasing flexibility in controlling thermal histories.

a

b

Fig. 4. Illustration of Voronoi scan strategy (a) object-dependent seed points placement (b) partially hatched layer filled with Voronoi fields

The concept of locally tailoring microstructure and mechanical properties through Voronoi-based partitioning has not yet been experimentally explored in the present work. The possibility of such tailoring emerged from a series of preliminary trials and publications[15, 2, 17, 13], and is therefore considered a promising direction for future investigations.

2.3. Experimental Methodology for Evaluation

Scan strategies can a ff ect virtually all aspects of component quality, including density and the associated formation of gas or keyhole pores, the occurrence of cracks, mechanical properties (e.g. tensile strength, fatigue resistance, residual stresses), microstructure, surface condition, and even the overall build success[9]. The primary objective of this project was therefore not an extensive optimization of typical manufacturing parameter, but rather optimization of conventional scan strategies and the the development of novel scan strategies. For comparability, conventional parameter sets were used as a baseline and were only minimally adjusted to enable straightforward transferability and benchmarking against established strategies. Specimens were fabricated on di ff erent PBF-LB / M, including an SLM ® 280 and an Evobeam SLaVAM 300. The main materials investigated were 316L (1.4404) stainless steel and the maraging tool steel Specialis ® (Rosswag GmbH, Germany).It is compositionally related to maraging steels M350[20], but modified by additional V and Al contents. To minimize the influence of machine-to-machine variation, material batch di ff erences, or operator e ff ects, di ff erent scan strategies were implemented within the same build job whenever possible, allowing direct comparison under identical boundary conditions. The reference parameters for the materials, along with the respective machinery on which they were utilized, are provided in Table 1. Table 1. Reference parameter for Specialis ® and 316L, the machine used for manufacturing and the form of investigation addressed in this paper with the material. Material layer height( µ m ) laser power ( W ) laser speed ( mm / s ) hatch distance ( µ m ) machine investigation

Specialis ® (1.6356)

50 60

250 368

850 925

100 140

SLM ® 280

fatigue

316L (1.4404)

SLaVAM 300 residual stress

In addition to demonstrator parts, which served to illustrate the general manufacturability of the individual strate gies, dedicated test specimens were produced for quantitative assessment. These included tensile specimens, fatigue